基于BP神经网络的城市轨道交通运营安全评价与设计

发布时间：2018-12-12 23:35

【摘要】：随着我国经济的快速发展,城市化水平也逐步提高,越来越多的人员涌入到城市中,城市人口的大规模发展势必造成城市生活中产生一系列问题,比如交通拥堵、环境污染等；为了解决城市地面交通拥堵问题,全国各地掀起了地铁建设的热潮,截至目前为止,全国已有20多个城市拥有地铁运营线路；同时由于地铁在运营过程中会聚集大量的客流,一旦发生运营事故,后果都比较严重,由此地铁运营安全问题逐渐成为人们关注的焦点,如果能对地铁运营系统做出有效的安全性评价,便可以做到事前控制,预防事故的发生。因此,辨识地铁车站运营安全的风险因素,实现地铁车站运营安全水平的有效评价,对提高整个地铁系统的运营安全水平具有重要意义。论文首先通过分析国内外相关领域研究现状,对地铁运营安全评价及人工神经网络理论进行了相应的研究；其次通过对国内外地铁运营事故进行统计分析,根据事故致因理论,确定了地铁运营事故的主要类型及导致地铁事故的主要原因,从系统论的观点出发,把影响地铁运营安全的影响因素划分为人、设备设施、环境和管理四个方面,在此基础上较为系统全面地构建了适合我国地铁运营安全的评价指标体系；接着构建了基于改进反向误差传播神经网络(Back Propagation; BP)的地铁运营安全系统评估模型,在计算方法上,用MATLAB的神经网络工具箱(Neural Network Toolbox; NNT)来进行网络设计和计算,由学习样本的训练和测试,使模型的误差达到预定的范围内,认为该模型在评价地铁运营安全性方面可行,并对新建成通车的杭州地铁运营安全状况进行了整体评价；文章最后将地铁运营安全评价模型设计成人机交互界面的形式,为更多不熟悉Matlab程序的用户提供了方便。
[Abstract]:With the rapid development of our country's economy, the level of urbanization is also gradually improving, more and more people pour into the city, the large-scale development of urban population will inevitably lead to a series of problems in urban life, such as traffic congestion, environmental pollution and so on. In order to solve the problem of urban ground traffic congestion, the upsurge of subway construction has been raised all over the country. Up to now, more than 20 cities in China have subway operating lines. At the same time, the subway will gather a large number of passengers in the operation process, once the operation accident, the consequences are more serious, so the subway operation safety issues gradually become the focus of attention. If we can make an effective safety evaluation of subway operation system, we can control beforehand and prevent accidents. Therefore, it is of great significance to identify the risk factors of subway station operation safety and to realize the effective evaluation of subway station operation safety level in order to improve the operation safety level of the whole subway system. Firstly, by analyzing the current research situation of related fields at home and abroad, the paper studies the safety evaluation of subway operation and the theory of artificial neural network. Secondly, through the statistical analysis of subway operation accidents at home and abroad, according to the theory of accident cause, the main types of subway operation accidents and the main causes of subway accidents are determined, from the point of view of system theory, The influencing factors of subway operation safety are divided into four aspects: human, equipment and facilities, environment and management. On this basis, the evaluation index system suitable for China's subway operation safety is constructed systematically and comprehensively. Then, the evaluation model of subway operation safety system based on improved back-propagation neural network (Back Propagation; BP) is constructed. In terms of calculation method, (Neural Network Toolbox; is used in MATLAB's neural network toolbox. NNT) is used to design and calculate the network. By training and testing the learning samples, the error of the model reaches a predetermined range. It is considered that the model is feasible in evaluating the safety of subway operation. The safety condition of the newly built Hangzhou Metro is evaluated as a whole. Finally, the paper designs the safety evaluation model of subway operation into the form of man-machine interface, which provides convenience for more users who are not familiar with Matlab program.
【学位授予单位】：大连交通大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U298

【相似文献】